Seizure count forecasting to aid diagnostic testing in epilepsy

Abstract

Objective: Epilepsy monitoring unit (EMU) admissions are critical for presurgical evaluation of drug-resistant epilepsy but may be nondiagnostic if an insufficient number of seizures are recorded. Seizure forecasting algorithms have shown promise for estimating the likelihood of seizures as a binary event in individual patients, but methods to predict how many seizures will occur remain elusive. Such methods could increase the diagnostic yield of EMU admissions and help patients mitigate seizure-related morbidity. Here, we evaluated the performance of a state-space method that uses prior seizure count data to predict future counts.

Methods: A Bayesian negative-binomial dynamic linear model (DLM) was developed to forecast daily electrographic seizure counts in 19 patients implanted with a responsive neurostimulation (RNS) device. Holdout validation was used to evaluate performance in predicting the number of electrographic seizures for forecast horizons ranging 1-7 days ahead.

Results: One-day-ahead prediction of the number of electrographic seizures using a negative-binomial DLM resulted in improvement over chance in 73.1% of time segments compared to a random chance forecaster and remained >50% for forecast horizons of up to 7 days. Superior performance (mean error = .99) was obtained in predicting the number of electrographic seizures in the next day compared to three traditional methods for count forecasting (integer-valued generalized autoregressive conditional heteroskedasticity model or INGARCH, 1.10; Croston, 1.06; generalized linear autoregressive moving average model or GLARMA, 2.00). Number of electrographic seizures in the preceding day and laterality of electrographic pattern detections had highest predictive value, with greater number of electrographic seizures and RNS magnet swipes in the preceding day associated with a higher number of electrographic seizures the next day.

Significance: This study demonstrates that DLMs can predict the number of electrographic seizures a patient will experience days in advance with above chance accuracy. This study represents an important step toward the translation of seizure forecasting methods into the optimization of EMU admissions.

Keywords: Bayesian inference; count forecasting; epilepsy monitoring unit; responsive neurostimulation; seizure prediction.

FIGURE 1

Flow diagram of patient selection criteria and data preprocessing steps. Steps 1–6 correspond to the data selection steps described in Section 2. LE, long episode; RNS, responsive neurostimulation

FIGURE 2

Sample forecasting results. (A) Combined results of model fitting and forecasting procedure for a sample patient. Long episodes (LEs) are shown on the log scale. (B) In-depth zoom into model fitting: expected number of long episodes, obtained as pointwise posterior means of negative binomial mean process from the training phase with the first 75% of the patient’s data. Data are shown with 95% credible bands (gray area). (C) In-depth zoom into seizure forecasting: one-step-ahead point forecasts for the remaining 25% of the data, with 95% posterior predictive intervals shown as the gray area

FIGURE 3

Heatmap of Pearson correlation matrix of responsive neurostimulation variables. Values are shown as Pearson correlation coefficients between pairs of variables. LE, long episode